Design Thinking for Startups
Rapid Qualitative Analysis & Synthesis: Using the Highlighter Add-on in Google Docs
Design thinking is a framework, both conceptual and tactical, that supports product innovation. At the core is a belief that people hold invaluable information about their experiences and needs, so it follows that the highest-value solutions emerge from end-user perspectives. The framework is typically broken down into five key stages, starting with inductive ethnographic data collection and qualitative synthesis during earlier stages (Empathize, Define, Ideate) and moving into more deductive, targeted questioning during later stages (Prototype, Test). The process is flexible and steps are repeated as learning progresses.
One indisputable fact about design thinking and ethnographic approaches in general is that high-quality analysis and synthesis takes significant time and energy. Simply moving from exploratory interviews to defining the problem often means culling many times through large amounts of content.
While resulting gains in product creativity and value are worth the effort, it is often difficult for startups to make the time and space to move from raw data to meaningful insights and actions.
This was the challenge I faced this year as the sole product person at Transmute — a five-member team urgently seeking deeper understanding of tangible business problems our technology might solve. My training is in ethnography and human-centered design, and I would have loved nothing more than to take a month sorting and externalizing (paper & post-it) the massive amount of information we collected over dozens of customer discovery interviews. However, we lacked dedicated office space, plus our team was partially distributed and needed to both see and contribute to the process. We also had to move quickly — testing and breaking insights and prototypes on a weekly, rather than a monthly basis.
I could not allow these constraints to stop us from effectively synthesizing our data and discovering valuable insights, so I had to come up with a fast and accessible alternative. Drawing from past experiences using qualitative analysis software like MAXQDA, I set out to establish a lean coding and synthesis system for our team using Google Docs and a free highlight tool add-on.
Here is how it works:
- Data Capture: I typed up notes from each discovery interview in an individual Google doc, using a standard template and naming convention (e.g. the interviewee initials and date of interview). I also saved highly related secondary data notes in a similar way, including quotes from videos and key take-aways from analyst reports.
- Data Consolidation: After the first set of interviews (in our case, twenty hour-long conversations) was documented, I pasted the text of each interview into one master Google doc, separating data by interviewee initials and date (hyperlinked to original notes).
- Coding Round 1 — Building Valuable Categories: While in the master Google doc, I went to Add-ons → Get add-ons and installed this free Highlight Tool. I read twice through all data, iteratively creating a new highlighter set as natural categories emerged. This first pass of highlighting was mostly about sense-making and sorting ideas, so I didn’t worry too much about exactly what the categories were called and instead trusted gut reactions. As a new category emerged, I created a color and label for it in the same set. About 80% of the notes received some highlight. This proved an effective digital approach to affinity diagramming. I then used the incredibly handy Extract Highlights → By Color feature. The result was a category-sorted export of all highlights to a new Google doc. Anyone familiar with sorting large qualitative data sets will understand why this is a jump-for-joy type moment!
4. Coding Round 2 — Synthesis aka “Panning for Gold”: I next removed all of the existing highlights from the beautifully sorted new document, and set to work coding a second time (with a new highlight set); this time looking for thought-provoking statements — notes that excited me and primed me for ideation or “How might we…” questions. I also used a different color to highlight points of tension or unresolved challenges. At this point our CEO also identified what she saw as important points, using the same highlight set which I shared with her. Upon completion I again used the Extract Highlights → By Color feature and ended up with a highly concise set of statements (down from 33 to about 3 pages).
5. Moving to Interpretation and Insight: I took the resulting document and started to build an ideation deck for our team. I crafted “How might we…?” statements (one per slide) out of the data. These slides were followed by supportive quotes from interviews.
6. Sharing with the team: The analysis, synthesis, and ideation process took about a week, and I was able to share each step along the way with our team. I presented the emergent insights deck to our Engineering team as a part of our roadmap planning, and we began breaking down ideas into user stories for prototyping and development. When questions emerged we were able to trace backwards through the sorted documents to gain more context. We also conducted follow-up interviews with some people to dive even deeper.
Our team found this process fit our needs (and time/resource constraints) in identifying the beachhead verticals and specific problem areas where our product drives real business value. We continue to iterate on this process as our product matures, and will continue to share that learning here. Please borrow from this approach and let us know how you make it work for your team!
